systematics of snakes of the dipsas oreas complex - BioOne

126 Bulletin Museum of Comparative Zoology, Vol. 158, No. 3
ferred to Dipsas gracilis are outside the
range of scale counts seen in Ecuadorian
specimens of D. gracilis, they are likewise
dissimilar to other species of Dipsas
known from western South America in
scutellation and/or color pattern. Dipsas
oreas is also known from the ‘‘valley of the
Río Quiroz’’ (Fig. 16, Fig. 23: locality 7),
the locality for MUSM 2700–01. However,
compared with characters of D. oreas
males (Table 1 for total sample, Table 3
for Peruvian populations), MUSM 2700–
01 have significantly greater numbers of
ventrals, subcaudals, and supralabials.
Moreover, the number of maxillary teeth
of MUSM 2700 (18) is much greater than
the number in a large sample of D. oreas
(12–14; Table 1). Collection of additional
material might permit a resolution of the
systematic status of these Peruvian populations,
particularly if new specimens of D.
gracilis from the geographic gap between
northern Peru and the vicinity of Guayaquil,
Ecuador, can be obtained. Biogeographically,
the presence of Dipsas gracilis
in lowland humid forests of Tumbes department
(MUSM 17589) was perhaps expected
on the basis of other geographic
and biological data from this region (e.g.,
Chapman, 1926; Koepcke, 1961), and
MUSM 17589 is well documented in the
context of a biological inventory (Wust,
1998a).
On the other hand, MUSM 2700–01 are
more problematic because they are not accompanied
by specific locality data or collector,
and there may be reason to question
their origin. Appropriate habitats do
not apparently exist presently in the lowlands
of the Río Quiroz valley, and Dipsas
gracilis is otherwise known only from lowland
localities. Moreover, D. latifasciata is
recorded from inter-Andean valleys immediately
east of the Río Quiroz (e.g., by
MCZ 17404; see comments below). Most
characteristics of the color pattern, scale
counts, body proportions, loreal scale pattern,
and maxillary tooth counts of MUSM
2700–01 are similar to those I recorded for
D. latifasciata from the Amazonian versant
of Peru and Ecuador (unpublished data).
One characteristic of the patterns of
MUSM 2700–01 is more typical of D. latifasciata
than of D. gracilis: The posterior
bands fail to encroach broadly onto the
ventral scutes or to meet midventrally. If,
in fact, MUSM 2700–01 were obtained
from east of the continental divide or if D.
latifasciata occurs on both versants of the
Andes in this region, then the identity of
these specimens should be reconsidered.
Dipsas latifasciatus and D. latifrontalis in
Eastern Ecuador and Peru
Dipsas latifasciata (Boulenger, 1913)
was described from the type locality ‘‘Upper
Marañon, eastern Peru,’’ and D. latifrontalis
(Boulenger, 1905) has the type locality
‘‘Aricaqua’’ [Venezuela]; Peters
(1960a: 110) identified the latter locality as
the town of that name on the Amazonian
slopes of the Andes in western Venezuela
south of Mérida. These names and their
applicability to populations in eastern Ecuador
require further study. Without examining
the holotype of either species, Peters
(1960a) assigned specimens from eastern
Ecuador to each. However, Peters’
(1960a) characterizations reveal similar
scutellation and few distinctions between
them, such as whether some or none of
the dorsal bands fail to meet midventrally,
as stated in Peters’ key (1960a: 33, couplet
35) or by the number of infralabial pairs
in contact behind the mental. The last
character is now known to be highly variable
within species of Dipsas; for examples,
see descriptions of the three species
discussed in detail in this report or Cadle
and Myers (2003) for D. nicholsi and D.
andiana. There is broad overlap in the scutellation
characters of these two taxa as
construed by Peters (1960a). This in itself
is not necessarily grounds for questioning
their validity, but in combination with the
other similarities noted by Peters and
some inconsistencies in his own discussions
(see below), it begs the question of
how these nominal taxa are distinguished